a thoughtful web.
Good ideas and conversation. No ads, no tracking.   Login or Take a Tour!
thundara's comments
activity:
thundara  ·  2440 days ago  ·  link  ·    ·  parent  ·  post: Trump Tells Aides He Has Decided to Remove Stephen Bannon

http://dailycaller.com/2017/08/17/bannons-time-is-up/

Daily caller, but it's Stone who's... a sight to behold:

    Not understanding that “personnel is policy”, Bannon refused to fight for any of his allies or those who helped get Trump elected. It’s as if Steve felt the grubby business of patronage was below him.

    Bannon delivered the Trump State Department into the hands of the Globalists.

    Bannon failed to push the President to use his most potent rhetorical and political weapon-Term Limits -in his current fight with the Congress. ... which sent the clear signal to Republicans in Congress that you can thwart the President and there are no consequences.

    Tough talking Anthony Scaramucci, whom I like and whom I think was treated rather shabbily, was, as I said, a political suicide bomber. He took out two RINOS: Preibus and Spicer
thundara  ·  2610 days ago  ·  link  ·    ·  parent  ·  post: Pubski: March 1, 2017

Truth. And yet it's just one control and a few experiments short of being a landmark paper in the field of Alzheimer's imo.

thundara  ·  2644 days ago  ·  link  ·    ·  parent  ·  post: Pubski: January 25, 2017

First thought: sweet and condensed semen

thundara  ·  2673 days ago  ·  link  ·    ·  parent  ·  post: Pubski: December 28, 2016

"Hopelash"? Was Cambridge the same for you? My friend who moved to Ithaca for grad school hated it by the first winter I saw him back home.

thundara  ·  2722 days ago  ·  link  ·    ·  parent  ·  post: President Donald Trump.

All three branches man...

thundara  ·  2784 days ago  ·  link  ·    ·  parent  ·  post: Apple Plug

And it plays fucking FLAC (Not a phone though)

thundara  ·  2785 days ago  ·  link  ·    ·  parent  ·  post: Pubski: September 7, 2016

Finished writing my thesis proposal, now in editing mode. Looking forward to a bit of time off after I give it. Booked a flight to Sweden.

thundara  ·  2794 days ago  ·  link  ·    ·  parent  ·  post: I Got Scammed By A Silicon Valley Startup

    ...and it goes on and on and on. I may go back and read this later but this seems a lot like the whiny rant of the one person stupid enough to take the job after the smart people ran screaming.

I submitted because it's a hilariously awful story of the scamming that goes on in the startup scene. I have a few friends who've shared similar stories (CEOs lying about funding, office leasers not being paid, etc), but this one takes the cake for absurdity.

The HN people tracked down the real company behind the story, and the CTO / co-founder at least had seemingly decent history of work. But otherwise red flag 6 and on-wards would make me start researching flights back home. Up to that point, I could see someone who was inexperienced or naive falling prey, thinking that was just the crazy world of startups. "hire fast and fire fast" isn't too far from "move fast and break things" in the list of SV-insane-utterances.

Then again, I would have turned the other direction at "* with machine learning / NLP".

With any luck, that CEO won't get any more money after a testimonial like this. Or maybe they'll go to jail for fraud.

Also, the author wasn't kidding about the spongebob facebook memes.

thundara  ·  2801 days ago  ·  link  ·    ·  parent  ·  post: What do you think of Dr Jill Stein and The Green Party?

    On your first two points, have you heard her explanation of how that would work?

Not encouraged to search myself, but happy to hear a summary. I'm highly doubtful of any plan that doesn't supplant debtors' payments by a gov't subsidy, which usually leads to a "tax-the-rich!" conclusion, which is handwavy.

    Healthy or not, if you're happy eating GMOs it doesn't mean everyone has to.

Hence understanding the labeling argument. Understand that would have a massive cost to the food processing pipeline as GMO vs. non-GMO would have to be tracked at every step along the pipeline, supermarkets would have to stock now three versions of food (organic, non-gmo, and gmo), etc. It's simply not as easy as "gluten-free" when it's not opt-in. The best-case scenario is you end up with another case of:

Which is an absolute joke back home.

    The point is that there are no long term studies, that's all she's asking for.

There are, 1, 2, 3, she just chooses to ignore them. The best anyone can point to to the contrary is Séralini, who is a scientific hack.

    Having a cautious approach to specific scientific discoveries is not being anti-science, it's being responsible.

When you're talking about crops that have been around for two decades, it's definitely anti-scientific to propose suddenly banning them without any evidence.

    And there's no need to patronize farmers abilities in order to defend GMOs. I'm sure they're smarter than you give them credit for.

It's not that they can't grow other crops, it's that there would be massive costs involved in suddenly forcing them to shift away.

thundara  ·  2821 days ago  ·  link  ·    ·  parent  ·  post: Washington State sues Comcast for $100m

    Removal of improper credit checks from the credit reports of more than 6,000 customers

Oh fuck yeah. I still have two marks against my credit score for the background checks they did on me when I signed up after moving to my current place.

thundara  ·  2862 days ago  ·  link  ·    ·  parent  ·  post: Pubski: June 22, 2016

Went to my first music festival last weekend, shit's wild. It felt like a hot, dusty version of white-peoples' heaven, filled with unabashed hedonism, drugs, thumping music, and sunburns as far as the eye could see. I got my month's worth of dancing in without re-injurying my knee and found a few new artists I enjoyed, including a glitch / edm / house / weeaboo artist, here's a song reminiscent of Daft Punk:

I also got 6x36 new mouse brain tissue samples last week for three more projects, which should round out my work schedule for the next year or so. Pretty much every major experiment I will be doing has now been planned. It should now just be a matter of a shitton of pipetting, coding, and learning a last few set of techniques like working with viruses and stem cells (Each of which could 1-2 months of training...).

I can't decide if I'm happier now that the rest of my aims / PhD proposal are starting to be set in stone or sad that the wild brainstorming will be shut down for a while.

thundara  ·  2869 days ago  ·  link  ·    ·  parent  ·  post: Pubski: June 15, 2016

Almost done reading every written work by DFW. Went through most of this audiobook this morning over 12 mini-preps, with the unabridged book in the mail some time later this month. Other than that, I'm finally taking some time off to visit Delaware this weekend.

thundara  ·  2888 days ago  ·  link  ·    ·  parent  ·  post: 2000 days. Thank you all.

2 hour flight, plus 1 hour bus, <$200, even on July 4th weekend. I can dig it.

thundara  ·  2889 days ago  ·  link  ·    ·  parent  ·  post: Could Alzheimer’s Stem From Infections?

Color me impressed that a paper coming out of Mass General / Harvard Medical that has absolutely no human data makes it into both Science Translational Medicine (A pretty good 2nd-tier journal) and NYTs.

Along with the fungi infection paper from last year, the main thing these guys have going for them is a somewhat out there theory, but a pretty minor amount of evidence. The link between Alzheimer's and Herpes / Diabetes is decently established (though why they are linked is still speculative). However I'd be pretty surprised if salmonella / fungus infections were causative in a majority of patients. That's the sort of thing that would pop out pretty quickly in epidemiological data...

I should note that I'm trying to replicate data from another paper by one of these authors right now and I'm pretty underwhelmed by their attention to detail.

thundara  ·  2918 days ago  ·  link  ·    ·  parent  ·  post: Pubski: April 27, 2016

Nothing quite saber-like, but not shown here is the reflectron which is a bit deeper inside the instrument and harder to extract:

thundara  ·  2981 days ago  ·  link  ·    ·  parent  ·  post: Pubski: February 24, 2016

I'm starting to appreciate why so few labs do proteomics (well).

When things go well, it's 1-2 weeks of work for often extremely useful data that can turn a year of measuring different things based on things the literature suggested to look at to a month of validating results on things no one would have bothered to investigate otherwise. After ~1 year, I have pretty much every protocol committed to memory, and I only occasionally have to look up exact volumes / concentrations.

When things go badly, it's months of wasted effort and up to thousands of dollars wasted in reagents alone, not including salary or instrument leasing costs. At this point, two post-docs, a tech, and I have switched out most of our reagents, tried new antibodies, and still the data coming out of each of our runs has been crap. It's a real crowbar in the bike wheels and has thrown off our entire lab's schedule for close to a month now.

In the mean time I'm catching up on my readings and trying to take my (one) class seriously. I have the growing suspicion that one of the professors may be the sort of person to take pleasure in ripping apart other labs' works. But I think it only makes me respect them more as I see how much crap makes it into even the top-tier journals of neuroscience.

In other news, I climbed my first V3 on Monday! It's nice to see some progression, since I'd definitely hit a plateau since coming back in January. Still have yet to move my body in any pattern even remotely resembling a dance though.

thundara  ·  3010 days ago  ·  link  ·    ·  parent  ·  post: The Villain of CRISPR

    Is there not any precedent as to what happens in cases like these? I feel like I've been hearing about patents on gene mapping, gene editing, even genes themselves, for a while now.

I'm not sure. The sequence of the original Cas9 gene found in S. pyogenes cannot be patented, as it is a naturally occurring DNA sequence. However, in this case, I believe the many modifications people have made to Cas9, as well as methods that use it to edit genes, constitute as patentable inventions.

As far as "who gets the patent", by law it's whoever was the "first to invent" this use of Cas9, but I'm not sure the exact details of how that will be argued in this case. Commenting to your friend that X could be used to do Y might be considered the first novel description of a process, but I don't think that's enough to hold water. But IANAPL.

    And an aside: how ground breaking is CRISPR? I've read a little bit about it, and listened to a Radiolab episode where a science writer answers questions that the hosts put to him, but I don't quite understand its scope. CRISPR is a gene editor. Had there been no editor before?

At its heart, Cas9 it a DNA nuclease. It cuts DNA. Plenty of enzymes cut DNA, but what makes it unique is that where it cuts is controllable. Again, other the cut sites of other enzymes are controllable, but what makes Cas9 particularly unique is that that control is specified by an RNA sequence.

TALENs and ZFNs are two other controllable-DNA-cutting-enzymes, but they both are targeted by the sequence of the protein. It's hard to predict in advance how different protein sequences will interact with DNA, so for ever new site you want to target, you must mutant, screen, and evolve a new protein. That can take months to years of work and is a barrier to use by most researchers.

In contrast, if you want to target a new sequence of DNA with Cas9/CRISPR, you simply find a sequence that is complementary to that sequence. In practice, there are still some caveats, but as a starting point, that gives you a lot more speed and power to new sequences or large sets of sequences.

----

The "editing" part of the equation comes from the fact that when DNA is cut, the cell has a number of repair pathways to reseal that break.

The one of the major ones is Non-Homologous End Joining (NHEJ), which just glues either end back together (sometimes inserting an extra bit of DNA in between). The other is Homology Directed Repair (HDR), which looks for DNA that is similar to the sequences around the cut site and copies that sequence over while sealing the break.

So in that latter case of HDR, if you also provide a short, similar sequence with the modification you want, the cell can repair that site, copying the new DNA, achieving that final goal of an "edit". It's not perfect, often the efficiency is rather low, but it's still proven to be a robust enough tool that people have rolled with it.

thundara  ·  3023 days ago  ·  link  ·    ·  parent  ·  post: Pubski: January 13, 2016

Big meeting tomorrow, doing last minute analysis, trying to improve my scripts to be more than "Sort by fold change and guess at the function of hits". It really amazes me how few good tools there are in the world of proteomics for my type of data. No GSEA for protein-based pathways, little consensus on clustering methods, poor gene ontology, incomplete interaction maps, no cell-type enrichment, very little annotation of even the most common protein modifications.

My research right now feels only a step above the average Reddit science discussion: find a hit, pubmed it, and speculate. The rest of my day will consist of learning more about self-organizing maps. Hopefully talking to some real neuroscientists will give more insight.

thundara  ·  3039 days ago  ·  link  ·    ·  parent  ·  post: 155th Weekly "Share Some Music You've Been Into Lately" Thread

Roommate's friend produces a bit of dnb and has reminded me of some old artists. Currently listening to Klute:

thundara  ·  3058 days ago  ·  link  ·    ·  parent  ·  post: Pubski: December 9, 2015

Gotta write a 10 page paper for tomorrow on ultrasound imaging, and then I'm done with classes for at least half a year. I feel like a world ahead of August-me, both in my research progress and neuroscience understanding. I have more ideas for next year than time, but I've just started taking an undergraduate under my wing, so maybe that will help.

It looks like the Morale Menagerie check-ins have stopped, but oddly enough, when I stopped thinking about them so much, things improved! Science and cooking have been going much better since my move (made some mulled wine last night), though drawing has been postponed to 2016.

thundara  ·  3086 days ago  ·  link  ·    ·  parent  ·  post: Pubski: November 11, 2015

This week: CLARITY, the technique for turning tissues transparent:

And a better image that is not my own:

thundara  ·  3093 days ago  ·  link  ·    ·  parent  ·  post: Pubski: November 4, 2015

Last week was non-human primates, but no lab visit as monkeys are apparently hyper-sensitive to TB and they didn't want to test us all. This week MEG:

thundara  ·  3114 days ago  ·  link  ·    ·  parent  ·  post: Apparently the Democratic Debate is tonight

Ah

    “But here’s where I do have concerns,” he continued. “There is a reason why Wall Street and all of corporate America likes immigration reform, and it is not, in my view, that they’re staying up nights worrying about undocumented workers in this country. What I think they are interested in is seeing a process by which we can bring low-wage labor of all levels into this country to depress wages for Americans, and I strongly disagree with that.”

    “There’s a very significant difference in scope of what the recent bill does compared to what that bill does,” Bernie said Thursday. “My concern about the bill that I voted against has to do with…that there was too much emphases on bringing low-wage workers into this country. What I want to see and what is better about the recent bill is that number one, there is a path towards citizenship which is absolutely essential. And second of all, that I was able to get a fairly significant amount of money into providing jobs for young people in this country.”

    At the USHCC event, Sanders also highlighted the more progressive parts of his immigration plan, which include comprehensive reform and a path toward citizenship for the roughly 11 million undocumented Americans.

    “Economically and morally, it is unacceptable that we have millions of workers who are living in the shadows,” he said. “Some of my Republican colleagues apparently think that the solution is I guess in the middle of the night to round up everybody and throw them out of the country. I think that anybody thinking those kinds of ideas is ugly beyond belief.”

I still wouldn't agree that that's inherently anti-Mexican. You could argue it's better for everyone if America changes its policies that depreciate the conditions elsewhere instead of trying to take their poor under its own roof (See: drug, trade, foreign policy). But I again maybe see the contrast with TNG's views

thundara  ·  3135 days ago  ·  link  ·    ·  parent  ·  post: Pubski: September 23, 2015

Back to things failing in lab. Classes have been neat though. Last week's topic: multi-photon microscopy:

Tomorrow is a demo on expansion microscopy and optogenetics. We ended up asking a bunch of questions about its developments and PI of the lab went into his unpublished slide deck and started showing updates on the technique for multiplexing this and getting better imaging on that. It's exciting stuff, and it looks like neuroscientists are going to have a prosperous next decade.

thundara  ·  3136 days ago  ·  link  ·    ·  parent  ·  post: Electrical Engineering vs. Computer Science (Toaster)

In the context of the modern age, they would now secure down the toaster by wrapping a RESToast API in a Docker container utilizing dynamic provisioning software to deploy micro toast applications (toastlets, if you will) that each perform a purely functional toasting operation leaving the hypertoaservisor in an immutable state.

Some modifications to the previous bootoastrap web-facing interface have been made to add a HTML5 material design that displays better on tablets.

thundara  ·  3136 days ago  ·  link  ·    ·  parent  ·  post: DIYBio Comes of Age

Meh, I think it's nice trend to see people re-creating the instruments most scientists treat somewhat as black-boxes, but having a thermocycler built with 21st century aesthetics doesn't really enable anything that you couldn't already do with a used $200 machine from ebay. And that's still pretty little.

Between reagents and a good collection of basic instruments, you're still looking at thousands to tens of thousands of dollars for even a minimal lab setup. It's not unfeasible to set up a garage lab, but it's also well outside most teens / early adults' budgets.

Some people have proposed microfluidics as a solution to lowering the cost by shrinking everything down, but that has yet to be proven practical as a general solution for biology. Most attempts to miniaturize existing protocols end up being their full-fledged research projects, and usually when you zoom you from the tiny plastic chip, you see hundred-thousand dollar microscopes / lasers / fluid control systems.

When it comes down to it, most lab instruments rely on pretty simple electronics / mechanics. A plate reader is a laser, a sensor, and a bunch of motors to move things around. A PCR machine is a heat pump, qPCR a heat pump plus laser and sensor. But the actual material selection, design, manufacturing, and quality control end up bumping up the price outside of the amateur's price range.

Also worth a read: Unicode: The Good, the Bad, and the (mostly) Ugly

Link

Sadly, the original website is more often down than not, but here are the unformatted slides:

https://dheeb.files.wordpress.com/2011/07/gbu.pdf

It has a good overview of the various feature support of different languages and problems that you can run into with things like regexes and passwords.

thundara  ·  3200 days ago  ·  link  ·    ·  parent  ·  post: An Honest Discussion About Transhumanism

    There seems to be an underlying assumption that humanity has not and therefore cannot make it. And, since humanity has seemingly destroyed any hope of biologically evolving as a species, our only hope is to inject our consciousness into machines. If you are at all interested in pop culture and what it projects onto and into people, you can see how we are being groomed for this idea as well.

Umm, it doesn't have to be this way. You're assuming the technology will get there, which it's not in danger of doing anytime soon, and ignoring the grayer area of: do I start implanting biochips into my body. The question there becomes less about morals and more risks and trade-offs. No technology will be perfect, and how many 9's of reliability do you need before you start putting that crazy retina chip in your body.

thundara  ·  3202 days ago  ·  link  ·    ·  parent  ·  post: The Misleading War on GMOs: The Food Is Safe. The Rhetoric Is Dangerous.

With regards to patents:

    Gene patents may claim the isolated natural sequences of genes, the use of a natural sequence for purposes such as diagnostic testing, or a natural sequence that has been altered by adding a promoter or other changes to make it more useful. In the United States, patents on genes have only been granted on isolated gene sequences with known functions, and these patents cannot be applied to the naturally occurring genes in humans or any other naturally occurring organism.[8]

https://en.wikipedia.org/wiki/Biological_patents_in_the_United_States#Gene_patents

    My other problem is that GMOs often are made for resistance to pesticides, and while that is great to ensure the harvest, the use of more and stronger pesticides is exactly what we shouldn't do, some pesticides are pretty harmless to humans, but those are not the ones GMOs are used to make crops tolerant against.

BT genes and RoundUp are both safe in the context of the average person's diet and safer for the environment than most other pesticides. Not that both don't encourage resistance, but that's a separate matter.

    I'm not convinced either by regulation or that there is sufficient and qualified oversight or structure to safeguard against blunders that simply wouldn't be possible with normal breeding

Most previous generation plants didn't really fare well outside the environment of a farm (since their ancestors evolved to work well on a farm, not outside of it). Horizontal gene transfer / cross-pollination are indeed risks, but we're not really spraying RoundUp on forests. To be fair though, horizontal gene transfer always going on between species. And in the case of many GMOs, there are other environmental benefits that partially offset the risks.

thundara  ·  3204 days ago  ·  link  ·    ·  parent  ·  post: Science Photos Tribe Too?